Impact of financial policies implemented in response to COVID-19 on global economy¶

Introduction¶

The Global Economy contracted sharply with the COVID-19 pandemic, leading heavily to a recession in most countries. Governments across the world were facing a health, social and economic crisis. Markets were hitting new lows; trades were on a halt and the population was in a panic while hospitals were overwhelmed. It was up to the government and financial institutions to roll out fiscal measures to combat the pandemic and protect the people.

Elliott Harris, the UN Chief Economist, and the Assistant Secretary-General for Economic Development stated that "The pace and strength of the recovery of the crisis not only hinges on the efficacy of public health measures in slowing the spread of the virus but also on the ability of countries to protect jobs and incomes, particularly of the most vulnerable members of our societies.”

Objective¶

Different types of fiscal policy measures taken across countries to tackle the pandemic is analysed in this notebook. Based on the types of policies taken by the best-performing countries^, policy measures will be recommended to the worst-performing countries^^.

To achieve this, data will be analysed to find the best and the worst-performing countries. This will narrow the search down to the top 10 and bottom 10 countries in terms of GDP percentage change. Analyses will majorly focus on each country's income level, policy implementation authority, and types of policy measures. This data would help recommend the type of policy measures the worst-performing countries should have taken to tackle the Great Lockdown better.

^Countries with the highest percentage change in 2020-2021 GDP as best-performing countries. ^^Countries with the worst percentage change in 2020-2021 GDP as worst-performing countries.

Dataset¶

The analysis was done by merging two datasets:

  1. "COVID-19 Finance Starter EDA" hosted by Bojan Tunguz in CSV format. The dataset consists of nearly 4000 rows with data on different policy measures taken by 157 countries.

  2. Annual GDP for every country in the world by the World Bank in CSV format.

The datasets contain the following attributes:

  1. COVID-19 Finance Starter EDA
  • Country Name
  • ID
  • Country ISO3
  • Income Level
  • Authority
  • Date
  • Level 1 policy measures
  • Level 2 policy measures
  • Level 3 policy measures
  • Details of the measure
  • Reference
  • Termination Date
  • Modification of Parent measure
  • Parent Measure
  1. Annual GDP by the World Bank
  • Country Name
  • Country Code
  • Indicator Name
  • Indicator Code
  • Year-wise GDP

Questions to answer:¶

  1. How did the GDP change in countries around the world before and during the pandemic?

  2. All countries went into losses but who was hit the worst?

  3. What steps were taken by the best performing countries to improve this economic situation?

  4. What policy measures taken by the better-performing countries can be suggested to the worst-performing countries?

  5. Are there low-income countries that performed relatively well in terms of increase in GDP?

Import files and libraries¶

In [1]:
#Importing required libraries for data analysis and visualization

import pandas as pd
import datetime as dt
import numpy as np

from matplotlib import pyplot as plt
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
In [2]:
covid_db = pd.read_csv('covid-fci-data.csv' , index_col='Country Name', encoding = "ISO-8859-1", skipinitialspace = True)
worldbank_db = pd.read_csv('worldbank_db.csv', index_col='Country Name', skiprows = 4, skipinitialspace = True)

Data cleaning¶

In [3]:
#Renaming White Space from Country ISO3 code
covid_db['Country ISO3']=covid_db['Country ISO3'].str.strip()
In [4]:
#Filtering the data as data related to pre, during and post covid years is required for analysis
gdp_db=worldbank_db[['Country Code','Indicator Name','Indicator Code','2018','2019','2020','2021']]
In [5]:
#Exception handling for chained assignment warning
pd.set_option("mode.chained_assignment",None)
In [6]:
#Getting observations with NaN values in gdp data
na_observations=gdp_db[gdp_db.isna().any(axis=1)]
#Dropping the observations with NaN values
gdp_db.dropna(inplace=True)
#Renaming the Country Code Column for further merging
gdp_db=gdp_db.rename(columns={'Country Code':'Country ISO3'})
In [7]:
#Resetting the index of Covid Database so that the Country Name is a column for cleaning
covid_db.reset_index(inplace=True)
#Filtering the covid database by removing the countries which have corresponding NaN values in GDP data
covid_db=covid_db[~covid_db['Country ISO3'].isin(na_observations.index)]

Merging of covid-fci-data with GDP dataset¶

The two databases are combined on the common Country ISO3 Code columns.

In [8]:
#Merging Covid Database and GDP database by inner joining on Country Name
merged_data=pd.merge(covid_db,gdp_db,how='inner',on='Country ISO3')
merged_data
Out[8]:
Country Name ID Country ISO3 Income Level Authority Date Level 1 policy measures Level 2 policy measures Level 3 policy measures Details of the measure Reference Termination Date Modification of Parent Measure Parent Measure Indicator Name Indicator Code 2018 2019 2020 2021
0 China 1 CHN Upper middle income SUP 2/1/20 Banking sector Operational continuity ... Blank ... Require banks and insurance companies to ensur... http://www.gov.cn/zhengce/zhengceku/2020-02/03... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
1 China 3 CHN Upper middle income CB 2/5/20 Banking sector Integrity ... Other integrity ... The PBC has issued Notices on AML/CFT requirem... NaN NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
2 China 9 CHN Upper middle income CB 2/17/20 Liquidity/funding Policy rate ... Blank ... Medium-term lending facility (MLF) rate cut by... https://www.reuters.com/article/us-china-econo... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
3 China 73 CHN Upper middle income SUP 3/3/20 Financial Markets/NBFI NBFI ... Insurance ... Commercial insurance companies will be support... https://www.china-briefing.com/news/china-covi... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
4 China 188 CHN Upper middle income CB 3/16/20 Liquidity/funding Liquidity (incl FX)/ELA ... Blank ... Reduce the RRR rate for certain qualified bank... NaN NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
3629 Niger 1334 NER Low income CB 4/3/20 Payment systems Easing regulatory requirements ... Blank ... Relaxation of requirements for opening electro... https://www.bceao.int/fr/communique-presse/com... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3630 Niger 1335 NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... Reduction of 50% on ATM fees on the GIM-UEMOA ... https://www.bceao.int/fr/communique-presse/com... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3631 Niger 1336 NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... Free mobile money transactions for amounts les... https://www.bceao.int/fr/communique-presse/com... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3632 Niger 1337 NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... \n Reduction of 50% on fees paid by the merch... https://www.bceao.int/fr/communique-presse/com... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3633 Niger 1338 NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... \n Free payment of water and electricity bill... https://www.bceao.int/fr/communique-presse/com... NaN No NaN GDP (current US$) NY.GDP.MKTP.CD 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10

3634 rows × 20 columns

Data Preparation¶

Only data relevant to analysis is kept; irrelevant columns are dropped from the merged dataset

In [9]:
#Defining a list with the relevant columns for analysis and cleaning
cols=['Country Name','Country ISO3','Income Level','Authority', 'Date', 'Level 1 policy measures','Level 2 policy measures', 'Level 3 policy measures', 'Details of the measure', 'Modification of Parent Measure', 'Termination Date', 'Indicator Name','2018', '2019','2020','2021']
merged_data=merged_data[cols]
merged_data
Out[9]:
Country Name Country ISO3 Income Level Authority Date Level 1 policy measures Level 2 policy measures Level 3 policy measures Details of the measure Modification of Parent Measure Termination Date Indicator Name 2018 2019 2020 2021
0 China CHN Upper middle income SUP 2/1/20 Banking sector Operational continuity ... Blank ... Require banks and insurance companies to ensur... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
1 China CHN Upper middle income CB 2/5/20 Banking sector Integrity ... Other integrity ... The PBC has issued Notices on AML/CFT requirem... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
2 China CHN Upper middle income CB 2/17/20 Liquidity/funding Policy rate ... Blank ... Medium-term lending facility (MLF) rate cut by... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
3 China CHN Upper middle income SUP 3/3/20 Financial Markets/NBFI NBFI ... Insurance ... Commercial insurance companies will be support... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
4 China CHN Upper middle income CB 3/16/20 Liquidity/funding Liquidity (incl FX)/ELA ... Blank ... Reduce the RRR rate for certain qualified bank... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
3629 Niger NER Low income CB 4/3/20 Payment systems Easing regulatory requirements ... Blank ... Relaxation of requirements for opening electro... No NaN GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3630 Niger NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... Reduction of 50% on ATM fees on the GIM-UEMOA ... No NaN GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3631 Niger NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... Free mobile money transactions for amounts les... No NaN GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3632 Niger NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... \n Reduction of 50% on fees paid by the merch... No NaN GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10
3633 Niger NER Low income CB 4/3/20 Payment systems Promoting and ensuring availability of digital... Blank ... \n Free payment of water and electricity bill... No NaN GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10

3634 rows × 16 columns

In [10]:
#renaming the columns.
merged_data.rename(columns={'Country ISO3':'Country Code','Date':'Policy Implementation Date','Details of the measure':' Policy Implementation Details','Modification of Parent Measure':'Policy Modifications','Termination Date': 'Policy Termination Date'}, inplace=True)
merged_data.head()
Out[10]:
Country Name Country Code Income Level Authority Policy Implementation Date Level 1 policy measures Level 2 policy measures Level 3 policy measures Policy Implementation Details Policy Modifications Policy Termination Date Indicator Name 2018 2019 2020 2021
0 China CHN Upper middle income SUP 2/1/20 Banking sector Operational continuity ... Blank ... Require banks and insurance companies to ensur... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
1 China CHN Upper middle income CB 2/5/20 Banking sector Integrity ... Other integrity ... The PBC has issued Notices on AML/CFT requirem... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
2 China CHN Upper middle income CB 2/17/20 Liquidity/funding Policy rate ... Blank ... Medium-term lending facility (MLF) rate cut by... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
3 China CHN Upper middle income SUP 3/3/20 Financial Markets/NBFI NBFI ... Insurance ... Commercial insurance companies will be support... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13
4 China CHN Upper middle income CB 3/16/20 Liquidity/funding Liquidity (incl FX)/ELA ... Blank ... Reduce the RRR rate for certain qualified bank... No NaN GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13

Data Cleaning¶

In [11]:
#Removing blank spaces in the country name
covid_db['Country Name']=covid_db['Country Name'].str.strip()

Replacing authority abbrevations with full form¶

  1. CB - Central Bank
  2. GOV - Government
  3. MoF - Ministry of Finance
  4. RES - Resolution Authority
  5. SUP - Supervisory Agencies
In [12]:
#remove white spaces from the column to rename
merged_data['Authority'] = merged_data['Authority'].str.strip()

#replacing blank values
merged_data['Authority'] = merged_data['Authority'].replace(r'^\s*$', np.nan, regex=True).fillna('No Information Available')

#Replacing abbreviations with full form for better understanding
merged_data['Authority'] = merged_data['Authority'].replace(['CB','GOV','MoF','RES','SUP'],['Central Bank','Goverment','Ministry Of Finance','Resolution Authority','Supervisory Agencies'])
In [13]:
# Converting into datetime format for consistency
merged_data['Policy Implementation Date'] = pd.to_datetime(merged_data["Policy Implementation Date"])

# Setting up a new date format ('%m-%d-%Y') for Date column
merged_data['Policy Implementation Date'] = merged_data['Policy Implementation Date'].dt.strftime('%m-%d-%Y')

# Making the date column formats consistent with eachother
merged_data['Policy Termination Date'] = pd.to_datetime(merged_data['Policy Termination Date']).dt.strftime('%m-%d-%Y')

#Convert into datetime format
merged_data['Policy Implementation Date'] = pd.to_datetime(merged_data['Policy Implementation Date'])
merged_data['Policy Termination Date'] = pd.to_datetime(merged_data['Policy Termination Date'])

#Extracting Policy implementation year and policy termination year for analysis
merged_data['Policy Implementation Year'] = merged_data['Policy Implementation Date'].dt.year.astype(int, errors='ignore')
merged_data['Policy Termination Year'] = merged_data['Policy Termination Date'].dt.year.astype(int, errors='ignore')

#convert year columns to int/string
merged_data['Policy Implementation Year'] = merged_data['Policy Implementation Year'].astype(str)
merged_data['Policy Termination Year'] = merged_data['Policy Termination Year'].astype(str)

#split on .
merged_data['Policy Implementation Year'] = merged_data['Policy Implementation Year'].str.split(pat='\.').str[0]
merged_data['Policy Termination Year'] = merged_data['Policy Termination Year'].str.split(pat='\.').str[0]

#Repositioning the implementation year and termination year
implementation_year = merged_data.pop('Policy Implementation Year')
termination_year = merged_data.pop('Policy Termination Year')

#Inserting the popped out year columns
merged_data.insert(5, 'Policy Implementation Year', implementation_year)
merged_data.insert(12, 'Policy Termination Year', termination_year)

#Creating a new column for Month from the date_time column
merged_data['Policy Termination Month & Year'] = merged_data['Policy Termination Date'].dt.strftime('%Y-%m')

#Creating a new column for Month from the date_time column
merged_data['Policy Implementation Month & Year'] = merged_data['Policy Implementation Date'].dt.strftime('%Y-%m')
In [14]:
#remove white spaces from the column
merged_data['Level 1 policy measures'] = merged_data['Level 1 policy measures'].str.strip()
#remove white spaces from the column
merged_data['Level 2 policy measures'] = merged_data['Level 2 policy measures'].str.strip()
In [15]:
# remove white spaces from the column
merged_data['Policy Modifications'].unique()
merged_data['Policy Modifications']=merged_data['Policy Modifications'].str.strip()
# remove white spaces from the column
merged_data['Policy Modifications'].unique()
merged_data['Policy Modifications']=merged_data['Policy Modifications'].replace('No','No Modifcations')

Data Analysis¶

  1. Determining best and worst performing countries from 2018 through 2021 by analysing GDP changes across the period
  2. Analysing Policy Implementation and Termination Months and Years due to pandemic
  3. Analysing Income Level of best and worst performing countries
  4. Analysing Authority of best and worst performing countries
  5. Analysing Level 1 Policy Measures of best and worst performing countries
  6. Analysing Level 2 Policy Measures of best and worst performing countries

Analysis 1: Determining best and worst performing countries from 2018 through 2021 by analysing GDP changes across the period¶

  • Defining a function to calculate Percentage Change of GDP and determing GDP changes for subsequent years (2018-2019, 2019-2020, 2020-2021)
  • Filter top 10 and bottom 10 countries for GDP changes across periods
  • Create an interactive mapbox chart for the GDP changes across the COVID pandemic period
In order to correctly analyze each country’s GDP growth, a metric has to be created to compare each country’s GDP growth regardless of the size or the GDP. For this, percentage change is calculated i.e. the GDP growth rate of each country’s GDP to run a comparison for the GDP change that occurred through the years of 2019 - 2021.¶
In [16]:
#Calculating GDP Changes across different years
def percentage_change(col1,col2):
    return ((col2 - col1) / col1) * 100

#Creating a new column for each year
merged_data['2018-2019 change'] = percentage_change(merged_data['2019'],merged_data['2018'])    
merged_data['2019-2020 change'] =  percentage_change(merged_data['2019'],merged_data['2020'])
merged_data['2020-2021 change'] =  percentage_change(merged_data['2020'],merged_data['2021'])
merged_data
Out[16]:
Country Name Country Code Income Level Authority Policy Implementation Date Policy Implementation Year Level 1 policy measures Level 2 policy measures Level 3 policy measures Policy Implementation Details ... Indicator Name 2018 2019 2020 2021 Policy Termination Month & Year Policy Implementation Month & Year 2018-2019 change 2019-2020 change 2020-2021 change
0 China CHN Upper middle income Supervisory Agencies 2020-02-01 2020 Banking sector Operational continuity Blank ... Require banks and insurance companies to ensur... ... GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13 NaN 2020-02 -2.696930 2.855309 20.741125
1 China CHN Upper middle income Central Bank 2020-02-05 2020 Banking sector Integrity Other integrity ... The PBC has issued Notices on AML/CFT requirem... ... GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13 NaN 2020-02 -2.696930 2.855309 20.741125
2 China CHN Upper middle income Central Bank 2020-02-17 2020 Liquidity/funding Policy rate Blank ... Medium-term lending facility (MLF) rate cut by... ... GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13 NaN 2020-02 -2.696930 2.855309 20.741125
3 China CHN Upper middle income Supervisory Agencies 2020-03-03 2020 Financial Markets/NBFI NBFI Insurance ... Commercial insurance companies will be support... ... GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13 NaN 2020-03 -2.696930 2.855309 20.741125
4 China CHN Upper middle income Central Bank 2020-03-16 2020 Liquidity/funding Liquidity (incl FX)/ELA Blank ... Reduce the RRR rate for certain qualified bank... ... GDP (current US$) 1.389482e+13 1.427994e+13 1.468767e+13 1.773406e+13 NaN 2020-03 -2.696930 2.855309 20.741125
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
3629 Niger NER Low income Central Bank 2020-04-03 2020 Payment systems Easing regulatory requirements Blank ... Relaxation of requirements for opening electro... ... GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10 NaN 2020-04 -0.834553 6.386607 8.802402
3630 Niger NER Low income Central Bank 2020-04-03 2020 Payment systems Promoting and ensuring availability of digital... Blank ... Reduction of 50% on ATM fees on the GIM-UEMOA ... ... GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10 NaN 2020-04 -0.834553 6.386607 8.802402
3631 Niger NER Low income Central Bank 2020-04-03 2020 Payment systems Promoting and ensuring availability of digital... Blank ... Free mobile money transactions for amounts les... ... GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10 NaN 2020-04 -0.834553 6.386607 8.802402
3632 Niger NER Low income Central Bank 2020-04-03 2020 Payment systems Promoting and ensuring availability of digital... Blank ... \n Reduction of 50% on fees paid by the merch... ... GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10 NaN 2020-04 -0.834553 6.386607 8.802402
3633 Niger NER Low income Central Bank 2020-04-03 2020 Payment systems Promoting and ensuring availability of digital... Blank ... \n Free payment of water and electricity bill... ... GDP (current US$) 1.280866e+10 1.291646e+10 1.374138e+10 1.495095e+10 NaN 2020-04 -0.834553 6.386607 8.802402

3634 rows × 23 columns

In [17]:
# Determing the countries with best and worst GDP changes from 2020 through 2021
best_gdp_2020_2021=merged_data.sort_values(by='2020-2021 change',ascending=False)
worst_gdp_2020_2021=merged_data.sort_values(by='2020-2021 change',ascending=True)
In [18]:
# Creating a dataframe for the top ten countries with best gdp changes for the year 2020-2021
best_gdp_2020_2021=best_gdp_2020_2021[['Country Name','2020-2021 change']]
best_gdp_2020_2021=best_gdp_2020_2021.drop_duplicates(ignore_index=True).head(10)
best_gdp_2020_2021.reset_index(drop=True,inplace=True)
best_gdp_2020_2021
Out[18]:
Country Name 2020-2021 change
0 Zimbabwe 45.241143
1 Haiti 44.362268
2 Guyana 35.420078
3 Angola 35.300713
4 Maldives 30.642999
5 Azerbaijan 27.941762
6 Ukraine 27.753971
7 Sudan 27.192137
8 Argentina 26.156060
9 Chile 25.454845
In [19]:
# Creating a dataframe for the top ten countries with worst gdp changes for the year 2020-2021
worst_gdp_2020_2021=worst_gdp_2020_2021[['Country Name','2020-2021 change']]
worst_gdp_2020_2021=worst_gdp_2020_2021.drop_duplicates(ignore_index=True).head(10)
worst_gdp_2020_2021.reset_index(drop=True,inplace=True)
worst_gdp_2020_2021
Out[19]:
Country Name 2020-2021 change
0 Lebanon -30.337649
1 Libya -19.955261
2 Myanmar -17.562908
3 Samoa -2.323932
4 Japan -2.037375
5 Jamaica -1.261140
6 Micronesia, Fed. Sts. -0.819079
7 Lao PDR -0.814739
8 Suriname -0.766788
9 St. Kitts and Nevis -0.484779

Lebanon is the worst-performing country with a drop of 30% from 2020 GDP to 2021 GDP. It is followed by Libya at a 20% drop, closely followed by Myanmar at a 17.5% decrease. (From subset worst_gdp_2020_2021)

COVID-19 contributed heavily to these losses, but it was not only poorly planned COVID-19 fiscal policies that lead to this drop. The external factors leading to this are:

  • Lebanon took a big hit on its economy with the Beirut port blast in 2020. This immediately led to a weakening local currency and losses in human life, food resources and trade. Source: Schlein, L. (2021) Economic fallout of Beirut Blast plunges millions into poverty, VOA. Voice of America (VOA News). Available at: https://www.voanews.com/a/middle-east_economic-fallout-beirut-blast-plunges-millions-poverty/6209118.html (Accessed: December 5, 2022).

  • Libya has been facing instability since the death of its former leader Muammar al-Qaddafi in October 2011. They faced deteriorating security issues with a large-displaced population. With the ongoing conflict severely impacting the country's medical system, they faced severe issues tackling COVID-19. Source: United Nations High Commissioner for Refugees. Libya: Humanitarian crisis worsening amid deepening conflict and covid-19 threat. UNHCR. https://www.unhcr.org/news/briefing/2020/4/5e86f2cc4/libya-humanitarian-crisis-worsening-amid-deepening-conflict-covid-19-threat.html. Accessed December 5, 2022.

  • Myanmar implemented COVID-19 restrictions early but internal unrest, inadequate testing capacity and an unprepared health system led to the worsening of its economy until it finally collapsed in 2022. Source: Welle (www.dw.com) D. Myanmar on brink of economic collapse one year after military coup | DW | 01.02.2022. DW.COM. https://www.dw.com/en/myanmar-on-brink-of-economic-collapse-one-year-after-military-coup/a-60621514 ‌ Traill AD Khaing Thandar Hnin, and Tom. Myanmar’s response to the COVID-19 pandemic. Brookings. Published December 1, 2020. https://www.brookings.edu/blog/future-development/2020/12/01/myanmars-response-to-the-covid-19-pandemic/ ‌

In [20]:
#Percent of GDP change for all countries from 2018 - 2021
gdp_change_percent = merged_data.melt(id_vars=['Country Name','Country Code'], 
                     value_vars=['2018-2019 change','2019-2020 change','2020-2021 change'], 
                     var_name='Year', value_name='GDP Change Percentage')

#Convert it into the appropriate datatype
gdp_change_percent['GDP Change Percentage'] = gdp_change_percent["GDP Change Percentage"].astype(float)

# Defining the mapbox plot
fig_map = px.choropleth(gdp_change_percent, locations='Country Code', hover_name="Country Name", 
                    color='GDP Change Percentage', 
                    scope='world', 
                    animation_frame ='Year', 
                    range_color=(-50,50),
                    #color_continuous_scale=px.colors.sequential.Viridis,
                    color_continuous_scale='Hot',
                    #color_continuous_scale=[(0, "red"), (1, "green")],  
                    projection="natural earth"
                    
                   )
# Adding Buttons 
fig_map.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 1000

fig_map.update_geos(visible=False, resolution=50, scope="world",
    showcountries=True, countrycolor="Black")
# Defining an appropriate margin
fig_map.update_layout(margin={"r":0,"t":25,"l":0,"b":0}, 
                  title="GDP % change from 2018 - 2021",
                  mapbox_style='open-street-map')
fig_map.show()
From the map, it is clear that there have been some drastic changes in the GDP for most countries around the globe. There is a decline in global economy during the COVID pandemic, and a slow and gradual rise in the aftermath. This can be attributed to the various fidcal policies for curtailing the economic effect of the pandemic. There are also some countries that performed quite well through the pandemic and some that did not.¶

Analysis 2: Analysing Policy Implementation and Termination Months and Years during pandemic¶

  • Determine the count of policies implemented and terminated for every month during pandemic
  • Create an interactive line chart with the number of policies per month and year (Implemented and Terminated)
In [21]:
# Creating a dataframe to analyse the policy termination date and the Number of Policies 
month_year_termination_data=merged_data[['Country Name','Policy Termination Month & Year']]
month_year_termination_data = month_year_termination_data.groupby('Policy Termination Month & Year').count()
month_year_termination_data.rename(columns={'Country Name':'Number of Policies'},inplace=True)
month_year_termination_data.reset_index(inplace=True)
In [22]:
# Creating a dataframe to analyse the policy Implementation date and the Number of Policies
month_year_implementation_data=merged_data[['Country Name','Policy Implementation Month & Year']]
# Using groupby() and count()
month_year_implementation_data = month_year_implementation_data.groupby('Policy Implementation Month & Year').count()
month_year_implementation_data.rename(columns={'Country Name':'Number of Policies'},inplace=True)
month_year_implementation_data.reset_index(inplace=True)
month_year_implementation_data
Out[22]:
Policy Implementation Month & Year Number of Policies
0 2020-02 31
1 2020-03 1208
2 2020-04 977
3 2020-05 404
4 2020-06 364
5 2020-07 175
6 2020-08 109
7 2020-09 128
8 2020-10 32
9 2020-11 25
10 2020-12 159
11 2021-01 11
12 2021-02 9
13 2021-03 1
In [23]:
# Create a line chart to display the Policy termination and implementation data
import plotly.graph_objects as go

termination_implementation = go.Figure()

termination_implementation.add_trace(go.Scatter(x=month_year_termination_data['Policy Termination Month & Year'], y=month_year_termination_data['Number of Policies'], name="Terminated Policies", mode='lines+markers'))
termination_implementation.add_trace(go.Scatter(x=month_year_implementation_data['Policy Implementation Month & Year'], y=month_year_implementation_data['Number of Policies'], name="Implemented Policies", mode="lines+markers"))

termination_implementation.update_layout(
    title="Number of Policies Implemented and Terminated all around the globe by month and year ", xaxis_title="Month Year", yaxis_title="Number of Policies"
)
termination_implementation.show()
While analyzing the implementation months and years, it can be observed that most of the economic policies were implemented at the start of the pandemic i.e. March 2020. There are also some countries who took the right measures looking at the sheer menace of the pandemic in the month of February 2020 itself. There were countries who implemented many policies at the end of the year. This may be due to the fact that the pandemic was supposedly on the decline.¶
Looking through the termination dates, it can be observed that some countries started terminating policies at the early stages from March 2020 through August 2020. This may be due to the fact that many of these policies were extremely short term. Also there may be a scenario where these policies might have failed and hence each country’s authoritative body wanted to abort these policies as soon as possible to save them from further trouble.¶

Analysis 3: Analysing Income Level of best and worst performing countries¶

  • Determining a dataframe with income level with best and worst performing countries in 2020-2021
  • Determining counts of Income Level among Best and worst performing countries
  • Create a bar chart for the counts of income level
In [24]:
# Creating a new dataframe to analyse income level data of best performing countries
income_best=merged_data[['Country Name','Income Level','2020','2021','2020-2021 change']]
income_best=income_best[income_best['Country Name'].isin(best_gdp_2020_2021['Country Name'])]
income_best=income_best.sort_values(by='2020-2021 change',ascending=False).drop_duplicates(ignore_index=True)
income_best
Out[24]:
Country Name Income Level 2020 2021 2020-2021 change
0 Zimbabwe Lower middle income 1.805117e+10 2.621773e+10 45.241143
1 Haiti Low income 1.450822e+10 2.094439e+10 44.362268
2 Guyana Upper middle income 5.471257e+09 7.409180e+09 35.420078
3 Angola Lower middle income 5.361907e+10 7.254699e+10 35.300713
4 Maldives Upper middle income 3.742770e+09 4.889667e+09 30.642999
5 Azerbaijan Upper middle income 4.269300e+10 5.462218e+10 27.941762
6 Ukraine Lower middle income 1.566179e+11 2.000855e+11 27.753971
7 Sudan Lower middle income 2.698756e+10 3.432606e+10 27.192137
8 Argentina Upper middle income 3.895910e+11 4.914927e+11 26.156060
9 Chile High income 2.527272e+11 3.170585e+11 25.454845
In [25]:
# Grouping by Income Level to calculate the frequency of income levels
income_best=income_best[['Country Name','Income Level']]
best_income_level = income_best.groupby('Income Level').count()
best_income_level.rename(columns={'Country Name':'Frequency of Income Levels'},inplace=True)
In [26]:
# Creating a new dataframe to analyse income level data of worst performing countries
income_worst=merged_data[['Country Name','Income Level','2020','2021','2020-2021 change']]
income_worst=income_worst[income_worst['Country Name'].isin(worst_gdp_2020_2021['Country Name'])]
income_worst=income_worst.sort_values(by='2020-2021 change',ascending=True).drop_duplicates(ignore_index=True)
income_worst
Out[26]:
Country Name Income Level 2020 2021 2020-2021 change
0 Lebanon Upper middle income 2.594892e+10 1.807662e+10 -30.337649
1 Libya Upper middle income 5.232022e+10 4.187958e+10 -19.955261
2 Myanmar Lower middle income 7.893026e+10 6.506781e+10 -17.562908
3 Samoa Upper middle income 8.071475e+08 7.883900e+08 -2.323932
4 Japan High income 5.040108e+12 4.937422e+12 -2.037375
5 Jamaica Upper middle income 1.381243e+10 1.363823e+10 -1.261140
6 Micronesia, Fed. Sts. Lower middle income 4.075159e+08 4.041780e+08 -0.819079
7 Lao PDR Lower middle income 1.898180e+10 1.882715e+10 -0.814739
8 Suriname Upper middle income 2.884248e+09 2.862132e+09 -0.766788
9 St. Kitts and Nevis High income 9.809059e+08 9.761507e+08 -0.484779
In [27]:
# Grouping by Income Level to calculate the frequency of income levels
income_worst=income_worst[['Country Name','Income Level']]
income_level_worst = income_worst.groupby('Income Level').count()
income_level_worst.rename(columns={'Country Name':'Frequency of Income Levels'},inplace=True)
In [28]:
# Create a plot to display the income levels of the best and worst performing countries
from plotly.subplots import make_subplots
import plotly.graph_objects as go

sub = make_subplots(rows=1, cols=2)

sub.add_trace(go.Bar(x=best_income_level.index, y=best_income_level['Frequency of Income Levels'],name="Frequency of Income Level types of Best performing countries"),
    row=1, col=1
)

sub.add_trace(
    go.Bar(x=income_level_worst.index, y=income_level_worst['Frequency of Income Levels'],name="Frequency of Income Level types of Worst performing countries"),            
    row=1, col=2
)

sub.update_layout(height=500, width=1400, title_text="Frequency of Income Level in Best and Worst performing countries in 2020-2021")
sub.show()
Haiti turned out to be a surprising find in the dataset. Despite being a low-income country, it had one of the highest positive changes in the GDP from 2020 to 2021. Haiti’s GDP improved by nearly 44% in 2021 from its GDP in 2020.¶
It is expected that high income countries would perform well during pandemic and implement better fiscal policies. However, from the graphs it can be observed that there are 2 high income countries that performed poorly - Japan and St. Kitts and Nevis. It is also surprising to see Haiti, a low income country, to have one of the highest GDP change of over 44%.¶

Analysis 3: Analysing Authority of best and worst performing countries¶

  • Determining a dataframe with Authority with best and worst performing countries in 2020-2021
  • Determining counts of Authority among Best and worst performing countries
  • Create a bar chart for the counts of Authority
In [29]:
# Creating a new dataframe to analyse authority data of best performing countries
authority_best=merged_data[['Country Name','Authority','2020','2021','2020-2021 change']]
authority_best=authority_best[authority_best['Country Name'].isin(best_gdp_2020_2021['Country Name'])]
authority_best.drop_duplicates(inplace=True)
authority_best.sort_values(by='2020-2021 change',ascending=False,inplace=True)
authority_best
Out[29]:
Country Name Authority 2020 2021 2020-2021 change
3126 Zimbabwe Supervisory Agencies 1.805117e+10 2.621773e+10 45.241143
3125 Zimbabwe Ministry Of Finance 1.805117e+10 2.621773e+10 45.241143
3113 Zimbabwe Goverment 1.805117e+10 2.621773e+10 45.241143
3111 Zimbabwe Central Bank 1.805117e+10 2.621773e+10 45.241143
2403 Haiti Central Bank 1.450822e+10 2.094439e+10 44.362268
459 Guyana Central Bank 5.471257e+09 7.409180e+09 35.420078
469 Guyana Goverment 5.471257e+09 7.409180e+09 35.420078
3152 Angola Goverment 5.361907e+10 7.254699e+10 35.300713
3151 Angola Ministry Of Finance 5.361907e+10 7.254699e+10 35.300713
3148 Angola Central Bank 5.361907e+10 7.254699e+10 35.300713
2415 Maldives Ministry Of Finance 3.742770e+09 4.889667e+09 30.642999
2411 Maldives Central Bank 3.742770e+09 4.889667e+09 30.642999
2414 Maldives Other 3.742770e+09 4.889667e+09 30.642999
893 Azerbaijan No Information Available 4.269300e+10 5.462218e+10 27.941762
869 Azerbaijan Other 4.269300e+10 5.462218e+10 27.941762
841 Azerbaijan Central Bank 4.269300e+10 5.462218e+10 27.941762
840 Azerbaijan Goverment 4.269300e+10 5.462218e+10 27.941762
2173 Ukraine Ministry Of Finance 1.566179e+11 2.000855e+11 27.753971
2163 Ukraine Central Bank 1.566179e+11 2.000855e+11 27.753971
3509 Sudan Central Bank 2.698756e+10 3.432606e+10 27.192137
778 Argentina Supervisory Agencies 3.895910e+11 4.914927e+11 26.156060
775 Argentina Goverment 3.895910e+11 4.914927e+11 26.156060
771 Argentina Other 3.895910e+11 4.914927e+11 26.156060
764 Argentina Central Bank 3.895910e+11 4.914927e+11 26.156060
2585 Chile Central Bank 2.527272e+11 3.170585e+11 25.454845
2591 Chile Other 2.527272e+11 3.170585e+11 25.454845
2595 Chile Goverment 2.527272e+11 3.170585e+11 25.454845
2596 Chile Supervisory Agencies 2.527272e+11 3.170585e+11 25.454845
2601 Chile Ministry Of Finance 2.527272e+11 3.170585e+11 25.454845
In [30]:
# Grouping by authority Level to calculate the frequency of authority levels
authority_best=authority_best[['Country Name','Authority']]
authority_best = authority_best.groupby('Authority').count()
authority_best.rename(columns={'Country Name':'Frequency of Authorities'},inplace=True)
authority_best.reset_index(inplace=True)
authority_best
Out[30]:
Authority Frequency of Authorities
0 Central Bank 10
1 Goverment 6
2 Ministry Of Finance 5
3 No Information Available 1
4 Other 4
5 Supervisory Agencies 3
In [31]:
# Creating a new dataframe to analyse authority data of worst performing countries
authority_worst=merged_data[['Country Name','Authority','2020','2021','2020-2021 change']]
authority_worst=authority_worst[authority_worst['Country Name'].isin(worst_gdp_2020_2021['Country Name'])]
authority_worst.drop_duplicates(inplace=True)
authority_worst.sort_values(by='2020-2021 change',ascending=True,inplace=True)
authority_worst
Out[31]:
Country Name Authority 2020 2021 2020-2021 change
3064 Lebanon Goverment 2.594892e+10 1.807662e+10 -30.337649
3062 Lebanon Central Bank 2.594892e+10 1.807662e+10 -30.337649
3137 Libya Central Bank 5.232022e+10 4.187958e+10 -19.955261
1628 Myanmar Central Bank 7.893026e+10 6.506781e+10 -17.562908
1629 Myanmar Goverment 7.893026e+10 6.506781e+10 -17.562908
3139 Samoa Goverment 8.071475e+08 7.883900e+08 -2.323932
3138 Samoa Central Bank 8.071475e+08 7.883900e+08 -2.323932
252 Japan Ministry Of Finance 5.040108e+12 4.937422e+12 -2.037375
274 Japan Goverment 5.040108e+12 4.937422e+12 -2.037375
255 Japan Central Bank 5.040108e+12 4.937422e+12 -2.037375
253 Japan Supervisory Agencies 5.040108e+12 4.937422e+12 -2.037375
3100 Jamaica Central Bank 1.381243e+10 1.363823e+10 -1.261140
3103 Jamaica Other 1.381243e+10 1.363823e+10 -1.261140
3472 Micronesia, Fed. Sts. Goverment 4.075159e+08 4.041780e+08 -0.819079
2769 Lao PDR Central Bank 1.898180e+10 1.882715e+10 -0.814739
3498 Suriname Central Bank 2.884248e+09 2.862132e+09 -0.766788
3497 Suriname Goverment 2.884248e+09 2.862132e+09 -0.766788
3550 St. Kitts and Nevis Central Bank 9.809059e+08 9.761507e+08 -0.484779
In [32]:
# Grouping by authority Level to calculate the frequency of authority levels
authority_worst=authority_worst[['Country Name','Authority']]
authority_worst = authority_worst.groupby('Authority').count()
authority_worst.rename(columns={'Country Name':'Frequency of Authorities'},inplace=True)
authority_worst.reset_index(inplace=True)
authority_worst
Out[32]:
Authority Frequency of Authorities
0 Central Bank 9
1 Goverment 6
2 Ministry Of Finance 1
3 Other 1
4 Supervisory Agencies 1
In [33]:
# Merging both the dataframes to create a single one for analysis
authority=pd.merge(authority_best,authority_worst,on='Authority')
authority.rename(columns={'Frequency of Authorities_x':'Best Performing Countries','Frequency of Authorities_y':'Worst Performing Countries'},inplace=True)
authority
Out[33]:
Authority Best Performing Countries Worst Performing Countries
0 Central Bank 10 9
1 Goverment 6 6
2 Ministry Of Finance 5 1
3 Other 4 1
4 Supervisory Agencies 3 1
In [34]:
# Creating a bar plot to display the different authority levels of both the Best and Worst Performing Countries
import plotly.express as px
# df = px.data.iris()
fig000 = px.bar(authority, x="Authority", y=["Best Performing Countries","Worst Performing Countries"],barmode='group',text_auto=True,
               labels={
                     "value": "Frequency of Authorities",
                   "variable":"Countries"
                 },
               color_discrete_sequence=["red", "blue"],
               )
fig000.update_xaxes(showgrid=False)
fig000.update_yaxes(showgrid=False)

fig000.update_layout(height=500, width=1400, title_text="Frequency of Authorities in Best and Worst performing countries in 2020-2021")
fig000.show()
It can be infered that both best and the worst performing countries had policies largely implemented by their central bank. This is followed by the government implemented policies, following by the Ministry of Finance. This helps identify which institutions make most fiscal policies for the countries.¶

Analysis 5: Analysing Level 1 Policy Measures of best and worst performing countries¶

  • Determining a dataframe with Level 1 Policy Measures with best and worst performing countries in 2020-2021
  • Determining counts of Level 1 Policy Measures among Best and worst performing countries
  • Create a scatter plot for the counts of Level 1 Policy Measures
In [35]:
# Creating a dataframe to analyze the level 1 policies of the best performing countries
level1_best=merged_data[['Country Name','Level 1 policy measures','2020','2021','2020-2021 change']]
level1_best=level1_best[level1_best['Country Name'].isin(best_gdp_2020_2021['Country Name'])]
level1_best.drop_duplicates(inplace=True)
level1_best.sort_values(by='2020-2021 change',ascending=False,inplace=True)
level1_best.sort_values(by='Level 1 policy measures',inplace=True)
print("Frequency of Level 1 Policy Measures in top 10 countries having best gdp change:")
level1_best_counts=level1_best['Level 1 policy measures'].value_counts()
level1_best_counts

Frequency of Level 1 Policy Measures in top 10 countries having best gdp change:
Out[35]:
Banking sector            10
Liquidity/funding          8
Payment systems            7
Financial Markets/NBFI     6
Insolvency                 1
Name: Level 1 policy measures, dtype: int64
In [36]:
# Creating a dataframe to analyze the level 1 policies of the worst performing countries
level1_worst=merged_data[['Country Name','Level 1 policy measures','2020','2021','2020-2021 change']]
level1_worst=level1_worst[level1_worst['Country Name'].isin(worst_gdp_2020_2021['Country Name'])]
level1_worst.drop_duplicates(inplace=True)
level1_worst.sort_values(by='2020-2021 change',ascending=False,inplace=True)
level1_worst.sort_values(by='Level 1 policy measures',inplace=True)
print("Frequency of Level 1 Policy Measures in bottom 10 countries having worst gdp change:")
level1_worst_counts=level1_worst['Level 1 policy measures'].value_counts()
level1_worst_counts

Frequency of Level 1 Policy Measures in bottom 10 countries having worst gdp change:
Out[36]:
Banking sector            10
Liquidity/funding          6
Payment systems            3
Financial Markets/NBFI     2
Insolvency                 1
Name: Level 1 policy measures, dtype: int64
In [37]:
# Create a scatter plot for analysing the Level 1 policy measures
import plotly.express as px

level1_plot_best = px.scatter(level1_best, x='Country Name', y='Level 1 policy measures',
             color='Level 1 policy measures',symbol='Level 1 policy measures')
level1_plot_best.update_traces(marker_size=10)

level1_plot_best.update_xaxes(showgrid=True)
level1_plot_best.update_yaxes(showgrid=True)

level1_plot_best.update_layout(height=500, width=1400, title_text="Level 1 Policy Measures Implemented by Best Performing Countries")

level1_plot_best.show()

level1_plot_worst = px.scatter(level1_worst, x='Country Name', y='Level 1 policy measures',
             color='Level 1 policy measures',symbol='Level 1 policy measures')
level1_plot_worst.update_traces(marker_size=10)

level1_plot_worst.update_xaxes(showgrid=True)
level1_plot_worst.update_yaxes(showgrid=True)

level1_plot_worst.update_layout(height=500, width=1400, title_text="Level 1 Policy Measures Implemented by Worst Performing Countries")

level1_plot_worst.show()
Negative economic growth (change in GDP) has become broad-based in many economies as they battle through the pandemic. The following observations can be made for the worst-performing countries, based on policy measures taken by the best-performing countries:¶
7 of the 10 best-performing countries aggressively promoted investments while all of the best-performing countries took measures to protect their banks and other financial institutions. Only 2 of the 10 worst-performing countries took measures to promote investments and 8 of the 10 countries did not take any measures to protect their banks and other financial institutions.¶
8 of the 10 best-performing countries took measures to increase liquidity in their markets. High liquidity helps countries and their people meet their short-term debts. Only 5 of the 10 worst-performing countries took measures to increase liquidity in the market.¶

Analysis 6: Analysing Level 2 Policy Measures of best and worst performing countries¶

  • Redefining Level 2 Policy Measures based on Policy implementation Details
  • Determining a dataframe with Level 2 Policy Measures with best and worst performing countries in 2020-2021
  • Determining counts of Level 2 Policy Measures among Best and worst performing countries
  • Create a bar chart for the counts of Level 2 Policy Measures
In [38]:
#list to recategorize data
support_borrowers = ['Other Payments','Support borrowers','Public debt management','Amending bankruptcy filing obligations','Policy rate']
market_movement = ['Operational continuity','Market functioning']
increasing_liquidity = ['Other liquidity','Consumer protection measures and ensuring availability and acceptance of cash','Easing regulatory requirements','Liquidity (incl FX)/ELA']
banking_protection = ['Crisis management','Prudential','NBFI']
digital_payments = ['Promoting and ensuring availability of digital payment mechanisms']
promoting_investments = ['Asset purchases','Integrity']


#add a new column
merged_data["Level 2 Recategorized policy measures"] = ''

#create a new column with recategorized data of Level 2 measures
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(support_borrowers), "Support Borrowers", merged_data["Level 2 Recategorized policy measures"])
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(market_movement), "Market Movement Measures", merged_data["Level 2 Recategorized policy measures"])
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(increasing_liquidity), "Increasing Liquidity", merged_data["Level 2 Recategorized policy measures"])
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(banking_protection), "Banking Protection", merged_data["Level 2 Recategorized policy measures"])
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(promoting_investments), 'Promoting Investments', merged_data["Level 2 Recategorized policy measures"])
merged_data["Level 2 Recategorized policy measures"] = np.where(merged_data["Level 2 policy measures"].isin(digital_payments), 'Promoted Digital Payments', merged_data["Level 2 Recategorized policy measures"])
In [39]:
# Creating a dataframe to analyze the level 2 policies of the best performing countries
level2_best=merged_data[['Country Name','Level 2 Recategorized policy measures','2020','2021','2020-2021 change']]
level2_best=level2_best[level2_best['Country Name'].isin(best_gdp_2020_2021['Country Name'])]
level2_best.drop_duplicates(inplace=True)
level2_best.sort_values(by='2020-2021 change',ascending=False,inplace=True)
level2_best=level2_best[['Country Name','Level 2 Recategorized policy measures']]
level2_best.drop_duplicates(inplace=True)
level2_best.sort_values(by='Level 2 Recategorized policy measures',ascending=False,inplace=True)
print("Frequency of Level 2 Policy Measures in top 10 countries having best gdp change:")
level2_best_counts=level2_best['Level 2 Recategorized policy measures'].value_counts()
level2_best_counts

Frequency of Level 2 Policy Measures in top 10 countries having best gdp change:
Out[39]:
Support Borrowers            10
Banking Protection            8
Promoting Investments         7
Increasing Liquidity          7
Promoted Digital Payments     6
Market Movement Measures      6
Name: Level 2 Recategorized policy measures, dtype: int64
In [40]:
# Creating a dataframe to analyze the level 2 policies of the worst performing countries
level2_worst=merged_data[['Country Name','Level 2 Recategorized policy measures','2020','2021','2020-2021 change']]
level2_worst=level2_worst[level2_worst['Country Name'].isin(worst_gdp_2020_2021['Country Name'])]
level2_worst.drop_duplicates(inplace=True)
level2_worst.sort_values(by='2020-2021 change',ascending=False,inplace=True)
level2_worst=level2_worst[['Country Name','Level 2 Recategorized policy measures']]
level2_worst.drop_duplicates(inplace=True)
level2_worst.sort_values(by='Level 2 Recategorized policy measures',ascending=False,inplace=True)
print("Frequency of Level 2 Policy Measures in bottom 10 countries having worst gdp change:")
level2_worst_counts=level2_worst['Level 2 Recategorized policy measures'].value_counts()
level2_worst_counts

Frequency of Level 2 Policy Measures in bottom 10 countries having worst gdp change:
Out[40]:
Support Borrowers            10
Banking Protection            6
Increasing Liquidity          5
Market Movement Measures      3
Promoting Investments         2
Promoted Digital Payments     2
Name: Level 2 Recategorized policy measures, dtype: int64
In [41]:
# Create a bar plot for analysing the level 2 policy measures
import plotly.express as px

level2_plot_best = px.bar(level2_best, x='Country Name', y='Level 2 Recategorized policy measures',
             color='Level 2 Recategorized policy measures')
level2_plot_best.update_layout(yaxis_title=None,height=500, width=1400, title_text="Level 2 Policy Measures Implemented by Best Performing Countries",yaxis_visible=False, yaxis_showticklabels=False)
# level2_plot

level2_plot_worst = px.bar(level2_worst, x='Country Name', y='Level 2 Recategorized policy measures',
             color='Level 2 Recategorized policy measures')
level2_plot_worst.update_layout(yaxis_title=None,height=500, width=1400, title_text="Level 2 Policy Measures Implemented by Worst Performing Countries",yaxis_visible=False, yaxis_showticklabels=False)
# level2_plot
level2_plot_best.show()
level2_plot_worst.show()
The following observations can be made on policies implemented by the worst performing countries as compared to the policies implemented by the best performing countries:¶
7 of the 10 best-performing countries actively promoted digital payments to prevent the spread of infection. 7 of the 10 worst-performing countries did not take measures to promote digital payments.¶
6 of the 10 best-performing countries took aggressive market movement measures to control the prices of goods in the short term (as seen from policy termination dates). Only 3 of the 10 worst-performing countries took market movement measures.¶
The highest positive change in GDP was in Zimbabwe with a 45% increase in GDP from 2020 to 2021. This was closely followed by Haiti at 44.36%. Guyana had the third-highest change at 35%. These countries took aggressive measures to protect their banks, support borrowers and increase liquidity in their market. Zimbabwe, the best performing country, took fiscal measures across all areas to ensure the stability of their financial institutions while supporting their population.¶

Conclusions & Recommendations¶

  • Government and financial institutions implementing temporary policy measures to support borrowers are necessary for most major economies to support the weakest segment of their population and ease inflation pressures. Every country primarily took measures to support borrowers and reduce debt across the nation. Even though supporting borrowers had good social benefits, it causes a decrease in value of currency due to increase in supply of money without any corresponding increase in goods and services available for purchase. Hence, it should be ensured that these measures are temporary to curb further inflation.

  • Only 2 of the 10 worst-performing countries took measures to promote investments and 8 of the 10 countries did not take any measures to protect their banks and other financial institutions. Maintaining the stability of the country's financial institutions significantly improves confidence in the economic situation of a country. Financial sectors control a country's stability, so protecting these institutions should be a priority to improve a country’s economy.

  • It is likely that the COVID-19 crisis is accelerating the shift towards digitalization. With 60% of best performing countries actively promoting digital payment mechanisms, it is speculated that the shift towards digitalization is accelerated because of this crisis. Countries that want to perform better on a longer run should start investing more in their digital economy. If they don't start soon, it could lead to further aggravation in net wage and unemployment of their people.